Accuracy evaluation of weather data generation and disaggregation methods at finer timescales

Availability of weather data at finer timescales such as hourly is vital in the application of dynamic physical and biological models. In this study, we have examined the suitability of various approaches (deterministic periodic versus stochastic) of disaggregating daily weather data into hourly data in the Cedar Creek watershed, TX, USA. We found the cosine function suitable to disaggregate daily maximum and minimum temperatures and wind speed data into respective hourly data. We also used a common logarithmic equation to compute vapor pressures from temperature data, and hence relative humidity (the ratio between actual and saturated vapor pressures multiplied by 100). Disaggregation following uniform distribution of daily rainfall over 24 h did not reproduce most statistical parameters computed from observed hourly rainfall data onsite. Conversely, both stochastic models formulated based on univariate (Hyetos) and multivariate (MuDRain) processes mimicked the measured hourly rainfall distributions very well. Overall, we found the MuDRain model superior, compared to other models to disaggregate daily rainfall data into hourly data.